How caffeine affects drug absorption
The use of caffeine is generally excluded among patients and healthy volunteers participating in phase I clinical trials. However, outside of a clinical setting, caffeine use is unrestricted. For many drugs, little consideration may be given to whether and how caffeine might influence the safety and efficacy of a prescribed medication. But perhaps it’s time to give this issue a bit more attention. Caffeine can influence drug absorption in three important ways:
Caffeine is classified as a weak inhibitor of cytochrome P450 CYP1A2 metabolic enzyme. 
Caffeine induces the production of gastric acid. 
Caffeine has been shown to increase the gastric emptying rate. 
Through CYP1A2 inhibition, caffeine interacts with a number of drugs. Among the most notable interactions are with tricyclic antidepressants, some antipsychotic drugs and warfarin. The interaction is through competition for available metabolic enzymes and the result is higher than normal plasma concentrations for both drug and caffeine.
The production of excess stomach acid can also exert a substantial influence on drug absorption through acid degradation of the drug, conversion of the drug substance to another form or by altering the drug’s rate of dissolution. The rate of acid catalyzed drug degradation is pH dependent and a relatively small change in pH can have a large impact on the amount of drug that survives gastric transit. For example, a drop of one pH unit was observed to reduce the amount of active midazolam by 75%.
Increased gastric acid generally results in faster dissolution of basic drugs. As illustrated in the image above, the effect of caffeine on the absorption of basic drugs can sometimes be quite dramatic. In this example, the same healthy volunteer absorbed a test drug 20 times faster when the drug was administered with caffeine. Changes of that magnitude to a drug’s absorption rate could have significant impact on the bioavailability of drugs that are either substrates for intestinal efflux transporters or undergo high first pass metabolism. In both cases, the effect would be an increase in systemic exposure for the drug.
Aspirin has been observed to deliver its analgesic effect faster and more effectively when co-formulated with caffeine. This is likely to be due to caffeine’s tendency to increase the gastric emptying rate. In the first place, faster gastric emptying makes the drug available for absorption sooner. In addition, aspirin undergoes rapid hydrolysis to salicylic and acetic acid in the stomach and the shorter gastric residence time reduces the amount of hydrolysis that occurs prior to absorption. Likewise, drugs that have been administered as salts of their parent form generally convert rapidly to the less soluble parent form, in the stomach. For these drugs, co-administration with caffeine may limit the time available for conversion, resulting in higher systemic exposure than if the drug were administered alone.
So what’s the point of all this? We who spend our time and talents developing new medicines can’t afford to work in a vacuum. We are trying to ensure that patients receive a safe and effective amount of drug with each dose. We want our medicines to perform exactly the same under real world conditions as they do in the tightly controlled clinical setting during phase 1. This is not possible unless we take into consideration the effect that individual consumption preferences can have on absorption of a drug in a real world setting. The effects of the various examples discussed above have ranged from reducing systemic exposure by half to doubling the expected exposure. Many of the effects could be managed or avoided through a better drug formulation strategy that takes into consideration the differences in how patients will take a medicine at home versus how the medicine is controlled in the clinic. It seems probable that a number of late stage clinical failures could be averted if this issue were more carefully considered and managed from the start of phase 1 drug development.
 British Journal of Pharmacology and Toxicology 2(3): 132-134, 2011
 Arch Intern Med. 1991;151:733-737